Acute respiratory distress syndrome (ARDS) is a major cause of morbidity and mortality. Of the many potential predisposing factors, sepsis and pneumonia represent the two main causes of ARDS. In spite of an increase in survival in recent years mortality in patients with ARDS is still estimated around 30 to 40%. In this context, development of effective preventive strategies in patients at high risk of development of ARDS is of paramount importance. Unfortunately, the results of studies evaluating prophylactic regimens for ARDS have been mostly disappointing. The gaseous molecule carbon monoxide (CO) has been traditionally viewed as a toxic metabolic and industrial waste. However, recent studies have demonstrated an important physiologic role of CO in many biological systems. Specifically, strong anti-inflammatory, anti-oxidant and anti-thrombotic effects of CO gas administration and heme oxygenase activation (the enzyme that generates endogenous CO gas) have been demonstrated in several animal models. Previous studies conducted in our department have demonstrated that bronchoscopic instillation of endotoxin (LPS) in healthy volunteers elicits a compartmentalized pulmonary inflammatory response, serving as an excellent model to evaluate interventions directed towards suppression of lung inflammation at its earliest stages. In the current single blinded, randomized, placebo controlled study, we are planning to evaluate the effects of inhaled carbon monoxide on local pulmonary inflammatory responses following endotoxin administration. Twenty healthy subjects, male or female, age 18 to 40 will undergo local endotoxin instillation, breath CO or room air through a mask for 6 hours, and then a repeat bronchoscopy with lavage will be done at 6 hours to assess the ability of CO to suppress local inflammation in the lung. We began enrolling subjects in May 2005 into the pilot study. To date, 9 subjects have been screened, 7 completed the pilot study (with CO inhalation for 3 hours or room air) and 1 did not been inclusion/exclusion criteria. The protocol is 05-CC-0012.